Detection and genetic characterization of feline bocavirus in Northeast China

Fecal samples were collected from private veterinary clinics and animal shelter centers in Shenyang, Jinzhou, Changchun, Jilin and Harbin in Northeast China between January 2016 and November 2017. Fresh feces were immediately collected using aseptic swab after defecation, and placed in RNase-free tubes. Samples collection was reviewed and approved by the Animal Care and Use Committee of Jilin Agricultural University, and was consented by client-owner. In total, 197 fecal samples were collected from 105 cats with diarrhea (the main clinical feature is acting depressed, anorexia, and the excretion of soft feces, watery or bloody stools.) and 92 healthy cats without any clinical symptoms and were examined in the present study. Approximately 1 g of fecal sample was homogenized in 5 ml of phosphate-buffered saline solution (PBS, pH = 7.4) using a vortex oscillator and then centrifuged at 10,000×g for 10 min at 4 °C to collect the supernatant. Each sample’s genome DNA was extracted from 250 μl of supernatant using the Viral DNA Extraction Kit I (OMEGA, China) according to the manufacturer’s instructions and then stored at − 70 °C until further use.

The genomic DNA extracted from fecal samples was analyzed for FBoVs using polymerase chain reaction (PCR) assays with three different primer pairs for FBoV-1, − 2 and − 3 as described in previous papers [9, 18, 19]. PCR was performed in 25 μl of reaction mixture containing 2 μl of template DNA, 12.5 μl of Premix Ex Taq (TAKARA, Japan), 1 μl of 10 pM forward and reverse primer and 8.5 μl of RNase free water using a thermocycler (BIOER, China). The primer sequences are shown in Table 1, and the amplification conditions were described previously. The sizes of the amplified DNA fragments were 133 bp for FBoV-1, 331 bp for FBoV-2 and 546 bp for FBoV-3. The amplified products were separated after electrophoresis on 1.5% agarose gels at 160 V for 20 min and visualized using a gel documentation system (Wealtec, USA). The prevalence of FBoVs and the distribution of the three genotypes was calculated, and the relationship between the prevalence of FBoVs and diarrhea symptoms in cats was analyzed with a univariable logistic regression model using the statistical program PASW Statistics 19.0. Probability (p) values < 0.05 were considered statistically significant.

To further determine the genotype of FBoVs detected in this study, 21 FBoV-positive samples were randomly selected for amplification and sequencing of partial NS1 gene. By aligning the nucleotide sequence of the NS1 gene among different FBoV genotype strains, two primer sets in the same region that could identify FBoV genotype and perform phylogenetic analysis, FBoV1-NS1F/FBoV1-NS1R targeted to a 705-bp region of the NS1 gene for FBoV-1 and FBoV2-NS1F/ FBoV2-NS1R to amplify a 963-bp fragment in NS1 gene for FBoV-2 (the primer sequences are shown in Table 1), were designed in the present study. We also optimized the reaction system and amplification conditions of these two primer sets in our study. Thirty nanograms of sample DNA was mixed with a total volume of 45 μl of reaction mixture containing 25 μl of Premix Ex Taq (TAKARA, Japan), 2 μl of 10 pM forward and reverse primers and 18 μl of RNase free water, and the amplification was performed with a thermocycler (BIOER, China) using the following conditions: initial denaturation at 94 °C for 5 min, followed by 35 cycles of 94 °C for 1 min, 62 °C for 1 min and 72 °C for 1 min, and a final extension at 72 °C for 7 min. PCR products were visualized on 1.0% agarose gels and purified using an AxyPrep DNA gel extraction kit (Axygen, China) according to the manufacturer’s instructions. The purified DNA was cloned in to a PMD-18 T vector, and positive clones were selected and sent to Sangon Biotech (Shanghai, China) for sequencing. To ensure sequence accuracy, each sample was sequenced using at least three positive clones, and each specific base was sequenced at least four times.

The full-length genomes of different FBoV genotypes from different regions were also determined using 5 pairs of specific primers for FBoV-1 and 5 pairs for FBoV-2, that we designed in this study. The nucleotide sequences of primers are shown in Table 1. The amplification was carried out in the same reaction mixture as described above, and the reaction conditions were as follows: predenaturation at 94 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 1 min, annealing at 60 °C for 1 min and extension at 72 °C for 90 s, and a final extension at 72 °C for 7 min. The PCR products were purified and cloned into PMD-18 T vectors according to the same procedures that we used in sequencing the partial NS1 gene. Positive clones were sent to Sangon Biotech (Shanghai, China) for sequencing. The nucleotide sequences were assembled using SeqMan program in the Lasergene 7.0 software (DNASTAR, USA), and the nearly complete FBoV genome sequences have been deposited in GenBank under accession numbers MH155946 – MH155951.

Partial NS1 gene sequences generated in this study were compared with those of other feline bocaviruses in GenBank using an online BLAST program (https://​blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi). The NS1 gene sequences alignment between FBoV strains identified in this study and representative bocavirus strains in the GenBank database (Additional file 1) were generated using ClustalW, and a phylogenetic tree based on the partial NS1 gene was constructed using the neighbor-joining (NJ) method with 1000 bootstrap replicates in MEGA 7.0 software.

For each complete genome that we obtained from FBoV-positive samples in the present study, the ORFs were predicted using two online programs: ORF finder (https://​www.​ncbi.​nlm.​nih.​gov/​orffinder/​) and Gene MarkS (http://​opal.​biology.​gatech.​edu/​GeneMark/​genemarks.​cgi). Phylogenetic tree of FBoVs identified in this study with different FBoV genotypes reference strains and other bocaviruses isolated from different species (Additional file 1) based on complete genomic nucleotide sequences was obtained using the neighbor-joining method, and 1000 replications were set for a bootstrap analysis. Maximum-likelihood (ML) phylogenetic analysis was also performed to confirm the results. SimPlot 3.5 software was used to determine the similarities of the FBoV genomes. The average percentage of nucleotide and amino acid identities was calculated using the MegAlign program. For phylogenetic analyses of NS1, NP1 and VP1, we constructed phylogenetic trees based on amino acid sequences. The possible recombination breakpoints were identified using the RDP method in the Recombination Detection Program (RDP) 4.0.

Out of 197 fecal samples, 51 (25.9%) samples were identified to be positive for FBoV using three pairs of specific primers for different FBoV genotypes. Of these FBoV-positive samples, 35 samples were positive for FBoV-1 and 12 were positive for FBoV-2. Interestingly, 4 samples were simultaneously identified as positive for FBoV-1 and FBoV-2 (Table 2). FBoV-1 was most prevalent in Northeast China. Detailed informations for the 51 FBoV-positive samples is shown in Additional file 2. We investigated the relationship between FBoV-infection and diarrhea symptoms, regions, season and residential environment (the sample’s source). The differences in the FBoV prevalence in different regions (19.2–29.4%) and different season (21.6–33.3%) were not significant. Interestingly, the two FBoV genotypes (FBoV-1 and -2) were present in all regions except for Harbin, while all samples that identified as co-infection of FBoV-1 and FBoV-2 were collected from Changchun (Fig. 1). Cats with diarrhea symptoms had a higher FBoV-positive rate (33.3%) than normal cats (17.4%). In the samples collected from animal shelter centers, the prevalence of FBoV was significantly higher than in samples from private veterinary clinics (Table 2). These results suggest that the FBoV infections were related to diarrhea symptoms and residential environments but were not associated with regions.

In this study, 21 FBoV-positive samples, including 13 FBoV-1-positive samples, 5 FBoV-2-positive samples and 3 coinfection samples, collected from different regions were randomly selected for sequencing of partial NS1 genes. Twenty-four nucleotide sequences of partial NS1 genes were obtained and have been been deposited in GenBank under accession numbers MH155928 – MH155951. These sequences were compared with different FBoV genotype reference strains and were compared to each other. Eight strains belonged to the FBoV-2 group and shared a high nucleotide and deduced amino acid identity with each other (98.6–100% nt identity and 99.1–100% aa identity) and with the FBoV-2 reference strain POR1 (98.0 - 99.6% nt identity and 98.3–99.6% aa identity). Sixteen strains belonged to the FBoV-1 group and showed 96.9–100% nt identity and 97.1–100% aa identity to each other, and 96.3–98.6% nt identity and 96.7–98.8% aa identity to the FBoV-1 representative strains HK797U, HRB2015-LDF and FBD2 identified in different countries. In the NJ phylogenetic tree based on the nucleotide sequence of partial NS1 gene, all of the feline bocaviruses could be grouped into three large branches: the FBoV-1, FBoV-2 and FBoV-3 branches. All of the FBoV-1-positive strains were placed in the FBoV-1 branch, and 8 FBoV-2-positive strains were placed in the FBoV-2 branch and displayed a closer relationship to the FBoV-2 strains from Japan (Fig. 2). To confirm our result, we also constructed a ML phylogenetic tree based on partial NS1 gene (Additional file 3). This second analysis also showed coincident grouping with the NJ phylogenetic analysis. Moreover, 16 FBoV-1 strains identified in this study and other FBoV-1 representative strains formed two major groups. Four of these 16 strains (16CC0803, 16CC1105, 17CC0302 and 17CC0508) and other reference strains from the USA, Hong Kong and Belgium were classified in group 2, whereas the other 12 strains clustered together forming group 1 and appeared to be more closely related to strain HRB2015-LDF from Northeast China in 2015 (Fig. 2).

A genomic analysis was performed using 6 nearly complete genomic sequences obtained from five FBoV-positive samples (included one mixed infection sample) in this study. The analytic results showed that the total genome length was identical for the same FBoV genotype strains: 5314 bp for all strains (16SY0602, 17HRB0511, 17CC0302 and 17CC0505(BoV1)) belonging to the FBoV-1 group and 5329 bp for the two other strains (16SY0701 and 17CC0505(BoV2)) belonging to the FBoV-2 group. Three ORFs (encoding NS1, NP1 and VP1) were predicted in each genome using Gene Marks and ORF Finder. The genetic organization of the full-length genome obtained in this study was coincident with that previously described for FBoV. The complete genome of FBoV-1 strains identified in the present study shared 96.8–98.7% nt identity and 94.9–97.0% aa identity with each other, and shared the closest nt and aa identities of 96.9–97.8% and 94.6–95.9% with two FBoV-1 reference strains (HK797U and HRB2015-LDF), respectively. The amino acid similarities of NS1, NP1 and VP1 genes between the identified FBoV-1 strains and reference strains were 96.9–97.8%, 95.9–97.3% and 94.9–96.9%, respectively. For the two Chinese FBoV-2 strains, the genome similarity was 98.9 and 98.6% at the nt level and aa level, respectively, when compared to each other. Greater identities (98.9–99.0% nt identity and 98.8–98.9% aa identity) were exhibited between the two Chinese FBoV-2 strains and the reference FBoV-2 strain (POR1); furthermore, for the NS1, NP1 and VP1 amino acid levels, the two FBoV-2 strains were 98.8–98.9%, 98.7–99.6% and 99.2–99.6% similar to the reference strain (Additional file 4). To further analyze the characteristics of the full-length genome of the six identified strains, a similarity plot analysis was performed using Simplot. Two prototype strains, HK797U for FBoV-1 and POR1 for FBoV-2, were used as query sequences to analyze the similarities of different genotype strains. 16SY0701 and 17CC0505(BoV2) displayed higher similarities than the other four FBoV-1 strains at the complete genome level when compared to query sequences. The four FBoV-1 strains identified in this study presented greater evolutionary variation in the NP1 and VP1/VP2 regions than in the NS region. However, higher similarities were exhibited in the NP1 and VP1/VP2 regions for the two Chinese FBoV-2 strains compared to each other and the query sequence (Fig. 3). We next investigated the recombination events in FBoV genome of 6 strains identified in this study and 5 different genotype strains deposited in GenBank using the RDP method. The analytical results indicated that no recombination breakpoints were found in any of the Chinese FBoV strains.

The NJ phylogenetic tree and ML phylogenetic tree based on the full-length genome sequences of the 6 strains identified in this study and other bocavirus sequences from various animals were performed. These six strains were more closely related to other FBoV strains than to other bocaviruses, and belonged to two different genotypes (Fig. 4 and Additional file 5). Then, we analyzed the phylogenetic relationship of different FBoV strains based on the deduced amino acid sequences in the NS1, NP1 and VP1 regions. In the phylogenetic trees, all FBoV-2 strains, including 16SY0701 and 17CC0505(BoV2), cluster together in three different regions. The FBoV-1 strains divided into three lineages in the VP1 region (Fig. 5b), whereas these lineages were not clearly observed in the NS1 and NP1 regions (Fig. 5a/5c). Particularly, 16SY0602 and 17CC0505(BoV1) clustered together in the NS1 and NP1 trees, but were placed into different branches in the VP1 region (Fig. 5).